Computer system, compute node, and data management method

Abstract

A computer system includes a NoSQL/SQL cluster and a distributed storage. In order to make storage target data redundant, a compute node in the NoSQL/SQL cluster instructs other compute nodes in the NoSQL/SQL cluster to write the storage target data into the distributed storage. As regards a file containing the storage target data, the compute node in the NoSQL/SQL cluster deduplicates data in storage apparatuses of a plurality of storage nodes in the distributed storage. The distributed storage performs erasure coding to store a file of the storage target data newly found to be duplicate in deduplication and to store a file of the storage target data not deduplicated in deduplication.

What is claimed is: 
     
         1 . A computer system comprising:
 a plurality of compute nodes; and   a plurality of storage nodes,   wherein the plurality of compute nodes form a database cluster, and the plurality of storage nodes form a distributed storage for managing data of the database cluster,   a compute node in the database cluster instructs the plurality of compute nodes in the database cluster to write storage target data into the distributed storage in order to make the storage target data redundant,   upon receipt of the instruction from the database cluster, as regards to a data unit including the storage target data designated by the instruction, the compute node deduplicates data in a storage apparatus to be managed by the plurality of storage nodes in the distributed storage, and   the distributed storage performs erasure coding to store a data unit of the storage target data that includes data newly found to be duplicate in deduplication and a data unit of the storage target data that includes data not deduplicated in deduplication.   
     
     
         2 . The computer system according to  claim 1 , 
 wherein the erasure coding has redundancy equal to or higher than the redundancy provided by making the data in the database cluster redundant.   
     
     
         3 . The computer system according to  claim 1 , 
 wherein, upon receipt of the instruction, the compute node identifies a record delimiter in the data units, and performs deduplication by regarding at least one record as a deduplication unit, the deduplication unit being the unit of deduplication.   
     
     
         4 . The computer system according to  claim 3 , 
 wherein, upon receipt of the instruction, the compute node calculates, in the deduplication, a hash value of the deduplication unit, and performs duplication determination based on the hash value.   
     
     
         5 . The computer system according to  claim 4 , 
 wherein, if a length of the deduplication unit is equal to or shorter than a predetermined length, upon receipt of the instruction, the compute node performs duplication determination by using a value of the deduplication unit without calculating the hash value of the deduplication unit.   
     
     
         6 . The computer system according to  claim 1 , 
 wherein one of the plurality of compute nodes and one of the plurality of storage nodes each include a physical computer.   
     
     
         7 . The computer system according to  claim 1 , 
 wherein the compute node performs the deduplication on the storage target data designated by the instruction before storing the storage target data in the distributed storage.   
     
     
         8 . A compute node that forms a database cluster together with other compute nodes and is connected to a distributed storage including a plurality of storage nodes, the compute node comprising:
 a processor,   wherein the processor 
 receives, from a compute node in the database cluster, an instruction for writing storage target data into the distributed storage in order to make the storage target data redundant, 
 as regards to the storage target data designated by the instruction, deduplicates data in storage apparatuses that are to be managed by the plurality of storage nodes in the distributed storage, and 
 sends an instruction to the distributed storage configured to store data units by erasure coding, the instruction directing the distributed storage to store a data unit of the storage target data that includes data newly found to be duplicate in deduplication and to store a data unit of the storage target data that includes data not deduplicated in deduplication. 
   
     
     
         9 . A data management method used by a compute node that forms a database cluster together with other compute nodes and is connected to a distributed storage including a plurality of storage nodes, the data management method comprising the steps of:
 receiving, from a compute node in the database cluster, an instruction for writing storage target data into the distributed storage in order to make the storage target data redundant;   as regards to the storage target data designated by the instruction, deduplicating data in storage apparatuses that are to be managed by the plurality of storage nodes in the distributed storage; and   sending an instruction to the distributed storage configured to store data units by erasure coding, the instruction directing the distributed storage to store a data unit of the storage target data that includes data newly found to be duplicate in deduplication and to store a data unit of the storage target data that includes data not deduplicated in deduplication.